Skip to content

[release/2.9][ROCm][inductor] Improved fast_tanh code generation #2804

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Nov 17, 2025
Merged

[release/2.9][ROCm][inductor] Improved fast_tanh code generation #2804

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
be87a96
On ROCm, always use fast_tanhf for triton codegen.
naromero77amd Nov 12, 2025
0b59f1c
Bump up Triton commit to support fast_tanhf.
naromero77amd Nov 13, 2025
427f9b0
Conditionalize fast_tanhf on triton_version.
naromero77amd Nov 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .ci/docker/ci_commit_pins/triton.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1 +1 @@
d704bc6e69c1a588c8edd3cbb67505d554ed65f6
5df9c723de8c23508773b07fe16dd34e4c444541
9 changes: 7 additions & 2 deletions torch/_inductor/codegen/triton.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,7 +26,7 @@
from torch._prims_common import is_integer_dtype
from torch.utils._ordered_set import OrderedSet
from torch.utils._sympy.functions import CeilDiv, FloorDiv, ModularIndexing
from torch.utils._triton import has_triton_package, has_triton_stable_tma_api
from torch.utils._triton import has_triton_package, has_triton_stable_tma_api, get_triton_version

from ...utils._sympy.symbol import free_symbol_is_type, prefix_str, symbol_is_type, SymT
from ...utils._sympy.value_ranges import ValueRanges
Expand Down Expand Up @@ -1315,7 +1315,12 @@ def tan(x):
@staticmethod
@maybe_upcast_float32()
def tanh(x):
return f"libdevice.fast_tanhf({x})"
if torch.version.hip and get_triton_version() > (3, 2):
# On ROCm, use fast_tanhf depending on Triton version
# Requires ROCm fork of Triton 3.3, 3.4, 3.5 or upstream Triton 3.6+
return f"libdevice.fast_tanhf({x})"
else:
return f"libdevice.tanh({x})"

@staticmethod
@maybe_upcast_float32()
Expand Down